MSSQL DataBase Backup without a specific table

问题:

I need to take a schedule backup without a specific table in sql. Because if I take a backup with that table it will take long time. I need to exclude one table from backup. Is it possible? Without that table all tables and data should be there in the database.

回答1:

You can setup a separate file group for this one table, apart from the PRIMARY file group. This will give you the ability to create a backup that omits your large table. Below is an example that steps out the process.

1) Create a new file group for your database.

USE [master]
GO
ALTER DATABASE [EXAMPLEDB] ADD FILEGROUP [EXAMPLEFG1]
GO

2) Create an identical table with a slightly different name on the new file group.

CREATE TABLE [dbo].[example]
(
    [e] [int] NOT NULL
)
ON [EXAMPLEFG1]

GO

3) Insert records from original table into new table.

4) Delete the old table and correct the name of the new table to match the name of the old table.

5) Backup PRIMARY which now excludes the table that is now on file group "EXAMPLEFG1".

BACKUP DATABASE EXAMPLE
   FILEGROUP = 'PRIMARY',
   TO DISK = '<Your Directory>'
GO

If you decide to do a backup of EXAMPLEFG1 simply change the FILEGROUP value to "EXAMPLEFG1" in the above query.

Check out this Microsoft site for more info on filegroup backups.

Hope this helps!

回答2:

Mike's answer is almost correct, except that you don't need to create another table to move the data.

According to this answer, which I tested myself, after creating a filegroup, you just need to move the clustered index to the other filegroup. If you don't have a clustered index, you can create one just to help you in this process. If you also have nonclustered indexes, move them too. After you finish, all data will have been moved between filegroups automatically.

So, if you want to exclude this big table from your backup routine, do the following:

1. Create a new file group.

   USE [master]  
   ALTER DATABASE [MyDatabase] ADD FILEGROUP [FG_MYBIGTABLE]
   

2. Add a new file to this file group.

   ALTER DATABASE [MyDatabase]
   ADD FILE
   (
       name = MyDatabase_MyBigTable,
       filename = 'C:\DB_Files\MyDatabase_MyBigTable.ndf',
       size = 1024MB,
       maxsize = unlimited,
       filegrowth = 100MB
   ) 
   TO FILEGROUP FG_MYBIGTABLE
   

3. Move the clustered index by creating another with DROP_EXISTING the old one.

   CREATE UNIQUE CLUSTERED INDEX CIX_MyBigTable
   ON MyDatabase.dbo.MyBigTable (ID)
   WITH DROP_EXISTING, ONLINE
   ON FG_MYBIGTABLE
   

4. Move the other indexes.

   CREATE NONCLUSTERED INDEX [IX_OtherTable] 
   ON MyDatabase.dbo.MyBigTable (OtherTable_ID)
   WITH DROP_EXISTING, ONLINE
   ON FG_MYBIGTABLE
   

5. Backup the database referencing the PRIMARY file group.

   BACKUP DATABASE MyDatabase
   FILEGROUP = 'PRIMARY',
   TO DISK = 'C:\Backups\MyBackup.bak'
   

回答3:

You can't exclude a table from a backup. That is kind of defines a backup, the whole thing, not part of it. You can however find some ways around this. One way you could do this is by creating a second database and copying all the data from your database except that table to this database and backup the copy. You could use replication to copy your data and exclude the table and backup the replicated database. The problem is that either of these might actually take longer due to the quantity of data in the rest of the database.